Reply to thread

Message

<blockquote data-quote="mm1352000" data-source="post: 1066894" data-attributes="member: 82144">What you've suggested is known as a ring buffer. This has been suggested previously though as far as I know nobody has attempted to implement it for MediaPortal. From a technical perspective it is more complex than the current approach, but perhaps &quot;cleaner&quot; or easier to configure/understand from a user perspective.Also, file size is limited on certain file system types. For example, FAT32 (still moderately common) has a maximum file size of 4 GB. That might be an issue for some people.My understanding is that timeshift buffer file space is pre-allocated to minimise overhead. In other words: writing to file that continuously grows in is more intensive for the storage subsystem than pre-allocating 256 MB and writing into that pre-allocated space.Don't quote me (I had no part in the original design choices - was long before my time - and I'm not an expert on file systems) but I'm not so sure that the inefficiency is minor. The HDD load/performance in part determines some key things like the maximum number of users and channel change speed. Therefore it makes sense to minimise the load on the HDD where possible.</blockquote>

[QUOTE="mm1352000, post: 1066894, member: 82144"] What you've suggested is known as a ring buffer. This has been suggested previously though as far as I know nobody has attempted to implement it for MediaPortal. From a technical perspective it is more complex than the current approach, but perhaps "cleaner" or easier to configure/understand from a user perspective. Also, file size is limited on certain file system types. For example, FAT32 (still moderately common) has a maximum file size of 4 GB. That might be an issue for some people. My understanding is that timeshift buffer file space is pre-allocated to minimise overhead. In other words: writing to file that continuously grows in is more intensive for the storage subsystem than pre-allocating 256 MB and writing into that pre-allocated space. Don't quote me (I had no part in the original design choices - was long before my time - and I'm not an expert on file systems) but I'm not so sure that the inefficiency is minor. The HDD load/performance in part determines some key things like the maximum number of users and channel change speed. Therefore it makes sense to minimise the load on the HDD where possible. [/QUOTE]

Verification